tf_estimate

1、 从网上下载人脸图像，构建人脸训练数据库和测试数据库，采用K-L变换进行特征脸提取，并实现人脸识别。通过K-L变换在人脸识别中的应用，加深对所学内容的理解和感性认识。(1, face image downloaded from the Internet, the construction of the training and test database face database, using K-L transform features face extraction, and realize the face recognition. Through k-l transform in face recognition application, deepen the understanding of the content of study and perceptual knowledge.)

利用MATLAB对利用液晶空间光调制器产生部分相干光进行关联成像（鬼成像）的过程进行仿真。(The process of correlated imaging using partial coherent light generated by liquid crystal space optical modulator is simulated.)

DSP toolbox for Matlab

its about how to calculate the distance by matlab

main function for retinal image viewer

在发送总功率限制下的，多用户MIMO块对角化（BD）预编码算法的和速率仿真。功率分配采用注水算法。(Send the total power limit, the multi-user MIMO block diagonalization (BD) precoding algorithm simulation and rate. Power allocation is water-filling algorithm.)

说明：  read from sp3 rinex in matlab

说明：  本程序实现了4PSK的调制与传输，计算了误码率等性能参数(This procedure achieved 4PSK modulation and transmission, calculated performance parameters such as bit error rate)

功率谱估计是利用有限长的数据估计信号的功率谱，广泛应用于各个领域。功率谱估计主要分为经典谱估计与现代谱估计。常用的经典谱估计方法有周期图法，相关法，周期图的改进法，常用的现代谱估计方法有最大熵谱估计，AR模型，MA模型，ARMA模型。经典谱估计适用于长序列的信号，其主要缺陷是描述功率谱波动的数字特征方差性能差，频率分辨率低，现代谱估计适用于短序列的信号，旨在改善谱估计的分辨率，并将其应用于实际地震资料的谱分析。 (Power spectrum estimation is the use of a finite length data to estimate the power spectrum of the signal, widely used in various fields. Power spectrum estimation is divided into classic and modern spectral estimation spectral estimation. Commonly used methods of classical periodogram spectrum estimation method, the relevant law, the cycle graph Improvement Act, commonly used in modern spectral estimation methods have maximum entropy spectral estimation, AR model, MA model, ARMA model. Classical spectral estimation applied to signal a long sequence, its main drawback is the fluctuation power spectrum describing digital signature variance performance is poor, the frequency resolution is low, modern spectral estimation applied to signal short sequences designed to improve spectral estimation resolution, and The analysis applied to real seismic data spectrum.)

基于小波的纹理特征，小波四层分解，特征向量含有均值、标准差、空间频率等，共13*4个特征。(Based on texture feature of wavelet, wavelet decomposition, four layers of characteristic vector containing the mean, standard deviation, such as spatial frequency, a total of 13* 4 characteristics.)